基于LS-SVM的军用航空发电机寿命预测研究  

Military aero-generator life prediction based on LS-SVM

作　　者：王莹玉[1] 崔建国[1,2] 赵云[2] 李威[2] 

机构地区：[1]沈阳航空航天大学自动化学院辽宁沈阳,110136 [2]沈阳飞机设计研究所,辽宁沈阳110035

出　　处：《沈阳航空工业学院学报》2010年第4期63-66,共4页Journal of Shenyang Institute of Aeronautical Engineering

基　　金：航空科学基金资助项目(项目编号:ID2008ZD54011);辽宁省教育厅科研基金资助项目(项目编号:ID2008544)

摘　　要：为解决航空发电机寿命有效预测的难题,以期实现视情维修,提升飞机的综合作战效能,提出一种基于LS-SVM(Least Squares Support Vector Machines)模型的航空发电机寿命预测方法。采用某型真实航空发电机,运用航空发电机专用寿命试验平台,对其进行加速寿命试验,获取输入转速、注油压力、负载、频率、电流、进油温度、出油温度、进口压力、出口压力等多种寿命相关参数,深入分析这些参数间的内在联系与寿命变化规律,设计LS-SVM寿命预测模型,并运用该寿命预测模型对航空发电机寿命表征参数进行预测。研究表明,所设计的LS-SVM寿命预测模型能较好实现对航空发电机的寿命预测效能,具有广阔的应用前景和较大的实际应用价值。To solve aero generator life prediction problem in order to realize condition maintenance,to improve the overall combat effectiveness of aircraft,a new methods,based on LS-SVM（least squares support vector machines,LS-SVM） model for life prediction of aero generators is proposed in this paper.Using a certain type of real aero-generator,and special life test platform for aero-generator,the accelerated life testing was conducted.And the input speed,fuel injection pressure,load,frequency,current,inlet oil temperature,outlet oil temperature,inlet pressure,outlet pressure and other relevant parameters is obtained,by in-depth analysis of the intrinsic link between these parameters and changes of lifecycle was carried out.A life prediction model is designed by LS-SVM,and model parameters of aero-generator to predict the life characterization is used.The results show that the LS-SVM model can realize the aero-generator life prediction and is of a wide application prospect and great practical value.

关 键 词：LS-SVM 航空发电机 寿命预测 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]